2262 J. Org. Chem.

1998, 63, 2262-2272

Effect of Ortho Substituents on the Direction of 1,2-Migrations in

the Rearrangement of 2-exo-Arylfenchyl Alcohols

Scott M. Starling, Simone C. Vonwiller,* and Joost N. H. Reek

School of Chemistry, University of Sydney, New South Wales 2006, Australia

Received November 4, 1997

A series of 2-exo-arylfenchyl alcohols (11a-k) was submitted to acid hydrolysis in EtOH/10 M HCl
(2:1-1:1, v/v) or with trifluoromethanesulfonic acid (TfOH) (1 equiv) in CHCl3 under varying
conditions. In all cases, initial formation of the cyclofenchene 12 took place, and after prolonged
treatment with acid the reaction proceeded along one of two pathways depending on the nature of
the aryl substituent. When the aryl substituent was o-NH2 11b or o-OH 11e, Wagner-Meerwein
rearrangement took place to give a carbocationic intermediate 9 that was trapped by the N or O
heteroatom to give (2R,4aR,9aR)-9-aza-2,4a-(10,10-dimethylmethano)-1,2,3,4,4a,9a-hexahydro-9a-
methyl-9H-fluorene (3b) and (3R,4aR,9bR)-3,9b-(10,10-dimethylmethano)-1,2,3,4,4a,9b-hexahydro-
4a-methyldibenzofuran (3c), respectively. In the case of 3c bearing an oxygen heteroatom,
equilibration to give the thermodynamic product (1R,4S,4aR,9bR)-1,2,3,4,4a,9b-hexahydro-1,4-
methano-1,4a,9b-trimethyldibenzofuran (4a), arising from Nametkin rearrangement, took place
via the intermediate 2-fenchyl cation. In contrast, when the aryl substituent was o-OCH3 11f,
direct conversion of the cyclofenchene to the Nametkin product 4a occurred with no detectable
prior formation of the Wagner-Meerwein product. In the case of o-tolylfenchyl alcohol 11k,
cyclofenchene formation was facile, but subsequent conversion to either the Wagner-Meerwein or
Nametkin products was highly disfavored. The results indicate that ortho substitution disfavors
Wagner-Meerwein rearrangement through adverse steric and electronic effects. However, when
the ortho substituent is NH2 or OH it is proposed that anchimeric assistance provides an
intermediate 15a that is stereoelectronically predisposed to Wagner-Meewein rearrangement.

Introduction positive charge onto the aryl ring causing a deviation

from nonclassical behavior,4 the subtle effects of intro-
The question of the nature of the norbornyl carbocation
ducing a 2-aryl substituent onto the fenchyl system have
1 represents one of the most thoroughly studied and
not previously been examined. Recently, we found that
controversial topics in the history of organic chemistry.
acid hydrolysis of (1′′S,2′′S,4′′R)-N-[4′-chloro-2-(2′′-meth-
The evolution of the concept of nonclassical carbocations
oxy-1′′,3′′,3′′-trimethylbicyclo[2.2.1]hept-2′′-yl)phenyl]-2,2-
and their later detection in superacid media can be
dimethylpropanamide to the corresponding amino alcohol
considered as milestones in the ensuing discussions
was followed by dehydration, Wagner-Meerwein rear-
promoted by Winstein, Brown, and Olah.1 The fenchyl
rangement, and trapping of the resulting carbocation
carbocation 2 has also been a subject of much interest,
with the free amino group to give an enantiomerically
and its most notable attribute is the large number of
pure hexahydroazafluorene 3 (Z ) NH, R ) H, R′ ) Cl)
rearrangements that it can undergo as indicated by the
(Scheme 1, path a). When the hydrolysis was carried out
characterization of five fenchenes (R-ζ) as well as
with protected phenols an alternative pathway, Namet-
cyclofenchene.2 While some work has been carried out
kin rearrangement to give 4, was followed, but the extent
to establish the relative energetics of these rearrange-
to which this occurred appeared to be intrinsically related
ments,2 the complexity of the system has precluded more
to the phenol substituent (Scheme 1, path b).5
extensive investigations and certainly limited synthetic
1,2-Migrations are susceptible to a number of influ-
application.3
ences among which stereoelectronic effects are the most
important.6 A particularly striking example is the rear-

While the effect of placing an aryl substituent onto C2
of the norbornyl carbocation serves to delocalize the
* To whom correspondence should be addressed. Tel.: +61 2 9351
4780. Fax: +61 2 9351 3329. E-mail: simone@chem.usyd.edu.au.
(1) For an excellent overview see: Schreiner, P. R.; Schleyer, P. v.
R.; Schaefer, H. F., III. J. Org. Chem. 1997, 62, 4216.
(2) Huang, E.; Ranganayakulu, K.; Sorensen, T. S. J. Am. Chem.
Soc. 1972, 94, 1779. Sorensen, T. S. Acc. Chem. Res. 1976, 9, 257.
(3) For example, see: Yuasa, Y.; Watanabe, T.; Nagakura, A.;
Tsuruta, H. King, G. A. III; Sweeny, J. G.; Iacobucci, G. A. Tetrahedron
1992, 48, 3473.
(4) Brown, H. C.; Takeuchi, K. J. Am. Chem. Soc. 1968, 90, 2693.
arnum, D. G.; Mehta, G. J. Am. Chem. Soc. 1969, 91, 3256. Brown,
H. C.; Takeuchi, K.; Ravindranathan, M. J. Am. Chem. Soc. 1977, 99,
2684. Substitution by electron-withdrawing groups increases the
tendency toward nonclassical stabilization: Farnum, D. G.; Wolf, A.
D. J. Am. Chem. Soc. 1974, 96, 5166.
(5) For preliminary results see: (a) Starling, S. M.; Vonwiller, S. C.
Tetrahedron Lett. 1997, 38, 2159. (b) Starling, S. M. Ph.D. Thesis,
University of Sydney, 1997.
(6) Schleyer, P. v. R.; Lam, L. K. M.; Raber, D. J.; Fry, J. L.;
McKervey, M. A.; Alford, J. R.; Cuddy, B. D.; Keizer, V. G.; Geluk, H.
W.; Schlatmann, J. L. M. A. J. Am. Chem. Soc. 1970, 92, 5246.
Majerski, Z.; Schleyer, P. v. R.; Wolf, A. P. J. Am. Chem. Soc. 1970,
92, 5731. Nickon, A.; Weglein, R. C. J. Am. Chem. Soc. 1975, 97, 1271.
Paquette, L. A.; Waykole, L.; Jendralla, H.; Cottrell, C. E. J. Am. Chem.
Soc. 1986, 108, 3739. Paquette, L. A.; Lanter, J. C.; Johnston, J. N.
J. Org. Chem. 1997, 62, 1702.

S0022-3263(97)02025-2 CCC: $15.00 © 1998 American Chemical Society

Published on Web 03/17/1998
Rearrangement of 2-exo-Arylfenchyl Alcohols
Table 1. Calculated (AM1) Heats of Formation ∆Hf (kcal mol-1) of Carbocation Intermediates and Corresponding
Rearrangement Products
Scheme 1
Scheme 2
rangement of [4.3.2]propellanes 5-7 (Scheme 2).7 In this
case, the concertedness or nonconcertedness of the mi-
gration and the orientation of the leaving group with
respect to the migrating bonds are the key issues.8
However, to our knowledge, there appear to be no
examples in which remote substituents are able so
dramatically to influence the direction of 1,2-migrations.
(7) Smith, A. B., III; Wexler, B. A.; Tu, C.-Y.; Konopelski, J. P. J.
Am. Chem. Soc. 1985, 107, 1308.
(8) See also: Kočovský, P.; Tureček, F.; Langer, V.; Podhlahová, J.;
Podlaha, J. J. Org. Chem. 1986, 51, 4888.
J. Org. Chem., Vol. 63, No. 7, 1998 2263
Therefore, in this paper we present a full account of our
results and a detailed analysis of the origin of the
substituent effect.
Results and Discussion
In line with literature results for the treatment of
fenchyl alcohol under superacid conditions,2 AM1 semiem-
pirical calculations on the aryl-substituted carbocations
8/8′ and the corresponding migrations show that the heat
of formation of the Wagner-Meerwein cationic precursor
9 is lower than that for the Nametkin precursor 10/10′
(Table 1). The relative energies of the final products were
calculated to follow the opposite trend with the Nametkin
products 4 being lower in energy than the Wagner-
Meerwein products 3 (Table 1). This suggested that if
the reactions are under thermodynamic control then
progressively harsher conditions should lead to a greater
proportion of Nametkin rearrangement in each case. To
delineate the factors affecting the regioselectivity of the
migrations, variously substituted 2-exo-aryl fenchyl al-
cohols 11a-k were prepared from fenchone and aryl-
lithium reagents9,10 and submitted to acid-catalyzed
rearrangement. The results are summarized in Table 2.
Rearrangement of the hydroxy pivalamide 11a in
refluxing ethanol/10 M HCl (1:1) led first to deprotection
and then loss of water to give the stable cyclofenchene
12a as a 73:27 mixture of atropisomers.11-13 Prolonged
heating in the presence of acid led to Wagner-Meerwein
(9) Exclusive formation of the 2-exo-aryl diastereomer in each case
was confirmed by NOE experiments. This contrasts with the behavior
of phenylmagnesium bromide, which reacts very sluggishly, requiring
the assistance of CeCl3, and predominantly from the endo side:
Dimitrov, V.; Bratovanov, S.; Simova, S.; Kostova, K. Tetrahedron Lett.
1994, 35, 6713. The need to activate fenchone with CeCl3 to nucleo-
philic attack by aryllithium reagents (see ref 10) is not supported by
our results.
(10) 11c (1R,2R,4S)-Enantiomer: Genov, M.; Kostova, K.; Dimitrov,
V. Tetrahedron: Asymmetry 1997, 8, 1869.
(11) Our initial assignment of the precursor to the hexahydroazaflu-
orene was incorrect: ref 5a.
2264 J. Org. Chem., Vol. 63, No. 7, 1998
rearrangement and trapping by the amino group to give
the final hexahydroazafluorene 3a in 63% yield after 48
h (Table 2, entry 1). Treatment of the amino alcohol 11b
(Table 2, entry 2) led to the same overall result except
that in the absence of a protecting group on nitrogen
rearrangement proceeded at a faster rate.14
Changes to the O-substituent had marked effects on
the ensuing 1,2-migrations of the hydroxy phenol deriva-
tives, and whereas o-hydroxyl (11e) and o-tetrahydropy-
ranyloxy (11d) groups caused varying degrees of Wagner-
(12) The formation of cyclofenchenes and related compounds is
relatively well-known in aprotic solvents: Bartlett, P. D.; Webster, E.
R.; Dills, C. E.; Richey, H. G., Jr. Justus Liebigs Ann. Chem. 1959,
623, 217; Ref: 13.
(13) Fry, J. L.; West, J. W. J. Org. Chem. 1981, 46, 2177.
(14) The acidic rearrangement conditions would appear to be
incompatible with the basic NH2 group and have a destabilizing effect
on carbocation formation. However, under aqueous conditions, the
pKa’s of protonated anilines are actually lower than those for carboxylic
acids; see: March, J. Advanced Organic Chemistry; John Wiley &
Sons: New York, 1992; p 251. Reichardt, C. Solvent Effects in Organic
Chemistry; VCH: Weinheim, 1990; pp 89-91. Arnett, E. M. J. Chem.
Educ. 1985, 62, 385. Thus, there will be an equilibrium mixture of
the unprotonated, monoprotonated, and diprotonated hydroxy anilines,
and it is expected that species i will be the only one significant for
carbocation formation and subsequent rearrangement. The effect of
protonation therefore is simply to decrease the rate of carbocation
formation.
Starling et al.
Figure 1. Three-dimensional representation of 3c with
selected NOE results.
Figure 2. Three-dimensional representation of 4a with
selected NOE results.
Meerwein and Nametkin rearrangement to take place
(Table 2, entries 4 and 5), o-methoxy groups (11f, 11g)
induced exclusive Nametkin rearrangement (Table 2,
entries 6 and 7). The structures of 3c (Figure 1) and 4a
(Figure 2) were unambiguously determined by NMR
spectroscopy through a combination of decoupling and
NOE experiments. The Wagner-Meerwein products
were characterized by a distinctive downfield doublet of
doublets of doublets due to H4-exo at 2.4-2.6 ppm while
the Nametkin products showed a narrow broad doublet
at ca. 2.2 ppm due to H4 at the bridgehead. On the basis
of the assumption that the energy barrier to Nametkin
rearrangement varied in each case due to different
degrees of carbocation stabilization, the reactions were
run at varying temperatures and reaction times. When
the ortho substituent was hydroxyl (Table 2, entry 5) it
was found that treatment with acid at 87 °C for 15 min
led to complete conversion into the Wagner-Meerwein
(3c) and Nametkin (4a) products in a ratio of 94:6. When
this product mixture was resubmitted to the same
reaction conditions for 7 h the ratio changed to 60:40,
indicating that the reaction was under thermodynamic
control, in agreement with the results in Table 1.
However, under conditions ranging from stirring at 20
°C to refluxing in acetic acid the preference for Nametkin
rearrangement for substrates bearing o-methoxy groups
(11f, 11g) did not alter, suggesting that somehow the
pathway leading to Wagner-Meerwein rearrangement
was blocked in these cases.15
To test whether the effect of the ortho substituent was
electronic or steric in nature, arylfenchyl alcohols bearing
a p-methoxy group (11h) (Table 2, entry 8) or an o-methyl
Rearrangement of 2-exo-Arylfenchyl Alcohols J. Org. Chem., Vol. 63, No. 7, 1998 2265

Table 2. Products Arising from Acid Hydrolysis of Substituted 2-exo-Arylfenchyl Alcohols

TfOH (1 equiv)/CHCl3
10 M HCl/ethanol (1:2 v/v) products (% yield),
entry 11 products (% yield), time/T (°C) time/T (°C)
1 a: RZ ) NH(CO)-t-Bu 12a (45%), 3a (29%), 24 h/refluxa
R1, R3 ) H 3a (63%), 48 h/refluxa
R2 ) Cl
2 b: RZ ) NH2 3b (68%), 24 h/refluxa
R1, R2, R3 ) H
3 c: RZ ) NMe2 no reaction, 6 h/refluxb
R1, R2, R3 ) H
4 d: RZ ) OTHP 3c/4a (95:5, 88%), 20 minc/refuxb
R1, R2, R3 ) H
5 e: RZ ) OH 3c/4a (97:3, 80%), 1.5 hc/60 °C (bath) 3c/4a (93:7, 78%), 5 hc/20 °C
R1, R2, R3 ) H 3c/4a (94:6, 97%), 15 minc/refluxb
3c/4a (60:40, 93%) 7.25 h/refluxb
6 f: RZ ) OCH3 12b (65%), 15 minc/85 °C (bath) 12b (74%), 4a (trace), 1 hc/20 °C
R1, R2, R3 ) H 4a (86%), 7.5 hc/85 °C (bath)
7 g: RZ ) OCH3 4b (88%), 1 hc/85 °C (bath)
R1 ) H 4b (33%), 4c (35%), 24 h/refluxd
R3 ) OCH3
8 h: RZ, R3 ) H 12c (57%), 1 h 40 minc/refluxb 13 (83%), 45 minc/20 °C
R1 ) OCH3
9 i: RZ ) CH2OH 14 (98%), 25 minc/refluxb
R1, R2, R3 ) H
10 j: RZ ) CH2OCH3 14 (99%), 1 h 50 minc/refluxb
R1, R2, R3 ) H
11 k: RZ ) CH3 12d (91%), 40 minc/refluxb 12d (24%), other (39%),f 24 h/reflux
R1, R2, R3 ) H 12d (84%),e other (16%), 3 d/refluxb
12d (64%),eother (36%), 24 h/refluxa
a EtOH/10 M HCl (1:1 v/v), 93 °C (internal temp). b 87 °C (internal temp). c Time required to effect complete reaction. d 17 M AcOH/
48% HBr (4:1, v/v). e Yields determined by analytical HPLC. f By weight.

group (11k) (Table 2, entry 11) were subjected to acid
hydrolysis conditions. However, in each case facile
formation of the cyclofenchenes 12c and 12d (as a 73:27
mixture of atropisomers) took place in refluxing 10 M
HCl/ethanol (1:2) within 40 min. This result did not
indicate whether the final outcome would be Nametkin
rearrangement or Wagner-Meerwein rearrangement as
the only other report of this type of tandem sequence
involving 11f by Fry and West,13 employing diethyl ether
as solvent, indicates that the cyclofenchene 12b is the
precursor to the Nametkin product 4a. It should be noted
that the product reported by Fry and West is erroneously
depicted as the Wagner-Meerwein product 3c.16 To
determine what the final products would be, the o-
methyl- (11k), o-methoxy- (11f), and p-methoxy-substi-
tuted (11h) adducts were subjected to acid-catalyzed
dehydration with an acid whose conjugate base is non-
nucleophilic, in an aprotic solvent. Thus, each was
treated with trifluoromethanesulfonic acid (TfOH) (1.0
equiv) in CHCl3 at 20 °C, and under these conditions the
reaction rates were somewhat increased over those
observed in 10 M HCl/ethanol (1:2). Surprisingly, while
the o-methoxy adduct led only to the cyclofenchene within
1 h (Table 2, entry 6), in agreement with the results of
Fry and West, the p-methoxy adduct underwent facile
conversion under the same conditions (entry 8) into the
alkene 13 arising from Wagner-Meerwein rearrange-
ment and proton loss. The o-methyl adduct 11k reacted
very sluggishly under these conditions and required
heating under reflux for 10 min to effect initial conversion
into the cyclofenchene 12d. Extended heating under
(15) Compounds 4b and 4c are of interest because of their structural
relationship to the cannabinoids. 4c can be considered as an analogue
of cannabielsoin although the latter shows no CNS activity: Uliss, D.
B.; Razdan, R. K.; Dalzell. H. C. J. Am. Chem. Soc. 1974, 96, 7372.
(16) The reported NMR data are consistent with our data for the
Nametkin product (see Experimental Section).
reflux in either medium resulted in slow conversion into
several products slightly more polar and considerably
more UV-active than the cyclofenchene (Table 2, entry
11). However, none of these was a Wagner-Meerwein
product as indicated by the absence of the characteristic
exo-methylene signals in the NMR spectrum. Treatment
of the o-(N,N-dimethylamino)-substituted adduct 11c
(Table 2, entry 3) in refluxing 10 M HCl/ethanol (1:2) did
not result in any change.
Thus, the effect of an ortho substituent is to inhibit
the Wagner-Meerwein migration and the result is that
elimination to give the cyclofenchene intermediate 12 is
the predominant outcome: the flanking methyl groups
make this the only mode of elimination possible. Forma-
tion of the cyclofenchene acts as a thermodynamic sink,
and reprotonation to provide the reverse reaction to the
fenchyl carbocation 8 is required to give access to the 1,2-
migration step (Scheme 3). In the presence of a proxi-
mate nucleophile, the less stable tertiary alkyl carboca-
tion 9 resulting from 1,2-migration is trapped to give the
final product 3. The possibility of nucleophilic ring
opening of the cyclopropane 12 is excluded on the grounds
that only a single regioisomeric hexahydroazafluorene is
formed when ZR ) NH2, and this corresponds to forma-
tion of the tertiary alkyl carbocation rather than the
secondary alkyl carbocation intermediate.
To accommodate the occurrence of the Nametkin
rearrangement, a competing pathway from the fenchyl
carbocation to the tertiary alkyl carbocation 10 resulting
from exo-methyl migration is set up. The exo-methyl
group is better aligned with the empty p-orbital than the
endo-methyl group as indicated by the calculated dihe-
dral angles for 3-Meexo-C3-C2-C1′ (Table 3),17 thereby
making this reaction stereoelectronically favored.18 Again,
trapping by a proximate nucleophile will give the final
Nametkin product 4.
2266 J. Org. Chem., Vol. 63, No. 7, 1998
Table 3. Comparison of Dihedral Angles (deg) of Energy-Minimized 2-Arylfenchyl Carbocations Calculated by AM1 and
MNDO Methods
aryl AM1
substituent C3-C2-C1′-C6′/3-Meexo-C3-C2-C1′ C3-C2-C1′-C6′/3-Meexo-C3-C2-C1′
R ) NH2, R′ ) H 19.2/68.6
R ) OH, R′ ) H 24.6/74.4
R ) OMe, R′ ) H 24.5/74.0
R ) R′ ) H 17.6/74.4
R ) H, R′ ) OMe 13.2/74.2
Scheme 3
The manner in which the ortho substituents inhibit
subsequent Wagner-Meerwein rearrangement is unclear
but is likely to be related to a steric effect of the fenchyl
methyl groups, which serves to cause the aryl ring to
deviate from being orthogonal to the empty p-orbital. The
effect is significant enough that even in the p-OMe
example where the ortho substituent is hydrogen the
reaction is slowed sufficiently in 10 M HCl/ethanol to
result in initial formation of the cyclofenchene. This
steric interaction would be expected to result in a more
localized cation through decreased orbital overlap and
decreased electron density at C219 and be particularly
significant for large substituents. While the large o-(N,N-
dimethylamino) substituent (11c) completely prevented
any reaction from occurring, this is likely to be due
mainly to the increased basicity of this group. AM1
results show that each carbocation possesses two energy
minima corresponding to conformations in which the aryl
ring adopts dihedral angles for C3-C2-C1′-C6′ ranging
from 13.2 to 24.6° (8) and 149.2 to 166.3° (8′) (Table 3).
Thus, according to these calculations, the aryl ring
prefers to be as close to orthogonal as possible. This
(17) For X-ray crystallographic data supporting the tendency for exo-
methyl migration in the Nametkin rearrangement, see: Moews, P. C.;
Knox, J. R.; Vaughan, W. R. J. Am. Chem. Soc. 1978, 100, 260.
Cameron, T. S.; Jochem, K.; Morris, D. G.; Maguire, J. Acta Crystallogr.
1994, C50, 2085.
(18) Brouwer, D. M.; Hogeveen, H. Recl. Trav. Chim. Pays-Bas 1970,
89, 211.
(19) Heagy, M. D.; Olah, G. A.; Prakash, G. K. S. J. Org. Chem.
1995, 60, 7355.
Starling et al.
MNDO AM1
C3-C2-C1′-C6′/3-Meexo-C3-C2-C1′
86.0/68.3 149.2/88.2
65.0/65.9 155.2/76.9
64.6/66.1 154.7/78.1
34.0/73.7 162.4/74.4
27.5/74.2 166.3/74.2
Table 4. Comparison of Heats of Formation ∆Hf (kcal
mol-1) and Dihedral Angles (deg) of Energy-Minimized
2-Arylnorbornyl Carbocations Calculated by AM1 and
MNDO Methods
aryl AM1 MNDO
substituent ∆Hf/C3-C2-C1′-C6′ ∆Hf/C3-C2-C1′-C6′
R ) H, R′ ) OMe 157.9/179.8 163.1/178.4
R ) R′ ) H 203.3/179.6 210.9/178.0
R ) OMe, R′ ) H 158.8/178.7 167.7/174.4
R ) NH2, R′ ) H 191.7/174.7 214.6/-139.5
condition appears to be favorable for Wagner-Meerwein
rearrangement based on the result of 11h, but this is
contrary to the idea that decreased electron density
through disruption of resonance stabilization or increased
electron demand provides an unstable carbocation that
is susceptible to neighboring group stabilization.4,19 On
the other hand, MNDO calculations suggest that for 8
(Z ) NH, R ) H) the aryl ring prefers to be in plane with
the empty p-orbital (dihedral ) 86.0°), a situation that
would preclude p-π overlap altogether. Resonance sta-
bilization is removed, leaving the inductive or field effects
of the aryl substituent to exert further effects on the
carbocation. As all of the substituents are inductively
electron-withdrawing,20 destabilization of the carbocation
must result, providing an impetus for migration to occur.
Thus, while both methods take into account electronic
effects, MNDO is much more sensitive to steric interac-
tions. This is supported by calculations performed on a
series of 2-arylnorbornyl carbocations in which both
methods afforded energy-minimized structures with C3-
C2-C1′-C6′ dihedral angles approaching 180° (Table 4).
It was only in the case of the larger o-NH2 substituent
that MNDO afforded a structure with a vastly different
dihedral angle of -139.5°. While more direct experi-
ments19 are required to assess the relative suitability of
each method to this problem, in reality it is expected that
a combination of steric and electronic factors will come
into play in determining the minimum energy conforma-
tions of these carbocations. Nevertheless, these observa-
tions do not fully address the question of regioselectivity
in the 1,2-migration.
The calculated rotational barriers for the cyclofenchene
atropisomers 12 (Table 5) correlate well with the ob-
served atropisomer ratios and give a good indication of
(20) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry Part
A: Structure and Mechanisms, Plenum Press: New York, 1990; p 201.
Rearrangement of 2-exo-Arylfenchyl Alcohols J. Org. Chem., Vol. 63, No. 7, 1998 2267

Table 5. Calculated (AM1) Rotational Barriers for Aryl-Substituted Cyclofenchene Atropisomers 12

∆Hf for lowest
aryl energy atropisomer rotational barrier atropisomer ratio
substituent (kcal mol-1) (kcal mol-1) (298 K)a
2′-CH3 47.1 17 73:27
2′-NH2, 5′-Cl 43.7 16.5 73:27
2′-OCH3 17.4 9.9 (lit.13 16.3) 52:48
2′-OH 10.6 9.8
4′-OCH3 14.5 4.7 unrestricted rotation
a Ratios measured by 1H NMR spectroscopy in CDCl3.

effective steric bulk of the ortho substituents. While the Scheme 4

OH and OMe groups would appear to be similar in size,
the results in Table 2 indicate that their effect on the
1,2-migration is quite different. Likewise, the effective
steric bulk of the NH2 group is similar to that of CH3,
and yet in the case of the former, Wagner-Meerwein
rearrangement proceeds much more readily.
Thus, it is clear that steric bulk alone is insufficient
to account for the diverging reaction pathways. However,
if the nucleophilicities of the ortho substituents are
considered then a correlation with tendency for Wagner-
Meerwein migration becomes apparent. This suggests Scheme 5
that an alternative pathway may be available that would
divert the fenchyl carbocation from the energy well of
cyclofenchene formation and assist the otherwise sluggish
Wagner-Meerwein rearrangement. It is proposed that
the o-NH2 and o-OH groups offer anchimeric assistance21
such that the fenchyl carbocation is transiently trapped
to give charged four-membered intermediates 15a and
15b (Scheme 3). Four-membered ring formation in
anchimerically assisted reactions is uncommon.22 Nev-
ertheless, when trapping occurs from the exo side the new
C-N or C-O bond in this intermediate 15a is now
antiperiplanar to the C1-C6 bond and a concerted
Wagner-Meerwein migration becomes stereoelectroni-
cally favored23 through sp3 alignment.6,7,24 In this case,
stabilization of the incipient cation by p-π overlap is
completely removed as the aryl ring is now in plane with
the developing empty p-orbital. The result is a species
that is more susceptible to stabilization by neighboring
groups, this being the C1-C6 bond in this case. While
the reactions involving 11a, 11b (Z ) NH) (Table 2,
entries 1 and 2) were difficult to monitor, those involving the reaction outcome is the same as for the phenol 11e.5
11e (ZR ) OH) (Table 2, entry 5) proceeded almost Trapping of the carbocation 8 from the endo side would
directly to the rearrangement product without significant lead to a species 15b with a lower tendency to undergo
buildup of the cyclofenchene.25 By comparison, in 10 M 1,2-rearrangement.
HCl/ethanol, formation of the Wagner-Meerwein product In light of this, the apparent anomaly with the result
is more facile for 11e (ZR ) OH) than for 11h (ZR ) H) of the p-anisylfenchyl alcohol 11h can best be reconciled
(Table 2, entries 5 and 8) despite the fact that OH is if the carbocation is considered to be sufficiently resonance-
larger than H. When the hydroxyl group is protected as stabilized and sterically unencumbered to allow revers-
a THP ether 11d hydrolysis to the phenol occurs im- ible trapping by water to compete with cyclofenchene
mediately in 10 M HCl/ethanol (Table 2, entry 4) and so formation to give an equilibrium mixture of the proto-
nated endo and exo alcohols (Scheme 4). The protonated
(21) Page, M. I. Chem. Soc. Rev. 1973, 2, 295. exo alcohol 16 will then meet the stereoelectronic re-
(22) Oxygen donor groups: Paquette, L. A.; Scott, M. K. J. Am.
Chem. Soc. 1972, 94, 6760. Gassman, P. G.; Marshall, J. L.; Mac-
quirements for concerted Wagner-Meerwein rearrange-
millan: J. G. J. Am. Chem. Soc. 1973, 95, 6319. Eliel, E. L.; Clawson, ment to occur.
L.; Knox, D. E. J. Org. Chem. 1985, 50, 2707. Sulfur donor groups: In the case of the OMe group the proposed exo four-
Eliel, E. L.; Knox, D. E. J. Am. Chem. Soc. 1985, 107, 2946.
(23) Winstein, S.; Trifan, D. S. J. Am. Chem. Soc. 1949, 71, 2953. membered intermediate 17a is unable to be formed or is
(24) For other relevant examples of apparently anchimerically too short-lived owing to steric interactions of the meth-
assisted Wagner-Meerwein rearrangements, see: (a) McCapra, F.; oxymethyl group with the methylene bridge and the
Beheshti, I. J. Chem. Soc., Chem. Commun. 1977, 517. (b) Kon-
dratenko, M.; El Hafa, H.; Gruselle, M.; Vaissermann, J.; Jaouen, G.; flanking methyls of the fenchyl skeleton (Scheme 5).
McGlinchey, M. J. J. Am. Chem. Soc. 1995, 117, 6907. While formation of the endo intermediate 17b is more
(25) By TLC a faint spot of intermediate polarity is evident likely to occur, this does not lead to Wagner-Meerwein
throughout the reactions and is likely to be the cyclofenchene, although
this was never formed in amounts sufficient for isolation and charac- rearrangement. The o-OMe group is likely to cause the
terization. aryl ring to deviate from being orthogonal to the empty
2268 J. Org. Chem., Vol. 63, No. 7, 1998
Figure 3. Three-dimensional representation of 14 with
selected NOE results.
p-orbital, thereby minimizing resonance stabilization,
and this together with the inductively electron-withdraw-
ing effect of the OMe group leads to carbocation desta-
bilization. Although C1-C6 bond migration to a cationic
center has a low tendency to occur in the 2-aryl fenchyl
system, exo-methyl migration takes place instead. When
two OMe groups are present as in 11g the rate of
Nametkin rearrangement increases over 7-fold (Table 2,
entry 7).
When the ZR group is the more bulky o-hydroxymethyl
(Table 2, entry 9), internal carbocation trapping is
relatively facile from the endo side and the result is
formation of the stable phthalan 14, which does not
undergo further reaction under the conditions used. The
stereochemistry of 14 was unambiguously determined by
NOE experiments (Figure 3). None of the product due
to trapping from the exo side was detected despite the
well-documented preference for nucleophilic attack of
fenchone from the exo side.9,10 Notably, when ZR is
o-methoxymethyl, internal carbocation trapping of 18 and
demethylation to give 14 are also facile from the endo
side, and again, this does not lead to 1,2-migration
(Scheme 5).
Thus, there is a narrow margin for the fenchyl skeleton
to tolerate the exo approach of o-aryl substituents, which
in these intramolecular cases have a restricted orienta-
tion and trajectory of approach.
Conclusion
In summary, acid hydrolysis of 2-exo-arylfenchyl alco-
hols affords the corresponding cyclofenchenes as the
kinetic products that upon prolonged treatment with acid
are converted into Wagner-Meerwein products via equili-
bration with the stabilized fenchyl carbocations. These
stabilized, sterically unhindered carbocations are pro-
posed to react with water to give 2-endo-arylfenchyl
alcohols that are stereoelectronically set up for Wagner-
Meerwein rearrangement. The presence of ortho sub-
stituents on the aryl ring hinders the Wagner-Meerwein
rearrangement through decreased resonance stabilization
of the carbocation and steric encumbrance to attack by
external nucleophiles and causes the reaction to enter
the alternative Nametkin manifold provided that a
cation-trapping group is present. However, when the
ortho substituent itself is nucleophilic the barrier to
Wagner-Meerwein rearrangement is overcome and this
is suggested to be due to internal trapping of the
carbocation from the exo side to give a reactive interme-
diate that is stereoelectronically predisposed to concerted
bond migration.
Starling et al.
Experimental Section
General Procedures. 1H and 13C NMR spectra were
recorded on Bruker AC-200F and AMX-400 spectrometers with
samples in CDCl3. Optical rotations were performed on a
Perkin-Elmer 241 polarimeter in cells with a path length of
0.5 dm. Flash chromatographic separations employed Merck
silica gel 60 (230-400 mesh). Analytical TLC was carried out
with Merck precoated aluminum plates coated with silica gel
60 F 254 (0.2 mm), and fractions were visualized by means of
UV light (278 nm) and developing the plate in vanillin/sulfuric
acid solution followed by heating. Elemental analyses were
performed at either the Research School of Chemistry, the
Australian National University, or the Campbell Microana-
lytical laboratory, University of Otago, New Zealand. Melting
points were determined on a Reichert micro melting point
apparatus and are uncorrected. All solvents and commercially
available reagents were purified by standard methods.26
Butyllithium as a solution in hexanes was standardized by
titration against 2,5-dimethoxybenzyl alcohol.27 Methyllith-
ium as a solution in diethyl ether was standardized by titration
against diphenylacetic acid.28
Molecular Modeling. All calculations were performed on
a Silicon Graphics IRIX workstation with SPARTAN 4.0.29 The
structures of the molecules listed in Table 1 were generated
with the Spartan program, minimized using the Sybyl force
field and further optimized with the semiempirical methods
AM1 or MNDO.
To calculate the rotational barriers of compounds listed in
Table 5, the dihedral angle (C3-C2-C1′-C2′) was forced to rotate
360° (in steps of 10°) using the coordinate driving option within
Spartan. The structures were minimized at each step using
the Sybyl force field and were further optimized with AM1.
The resulting energy-minimized structures were further mini-
mized by using AM1 without constraints, and the energy
barriers were calculated by subtracting the minima from the
maxima.
Preparation of Arylfenchyl Alcohols. General Method.
The lithiated benzene, prepared as described for individual
examples, was treated with (+)-fenchone (1 equiv) and added
dropwise as a neat liquid, and then the whole was stirred for
the stated time and at the stated temperature. The reaction
was quenched with saturated ammonium chloride solution,
and the product was extracted into ether. The combined
organic layers were dried (Na2SO4), and the solvent was
removed under reduced pressure. The pure product was
isolated as described for individual examples.
(1′′S,2′′S,4′′R)-N-[4′-Chloro-2-(2′′-hydroxy-1′′,3′′,3′′-
trimethylbicyclo[2.2.1]hept-2′′-yl)phenyl]-2,2-dimethyl-
propanamide (11a). Butyllithium (1.6 M, 3.90 mL, 6.24
mmol) was added to a solution of 4-chloropivanilide (656 mg,
3.10 mmol) in THF (10 mL) cooled to -5 °C to give a yellow
solution that was stirred at 0 °C for 2 h. The resulting solution
was treated with fenchone at room temperature for 3 h. The
crude product was purified by flash chromatography (ethyl
acetate/hexanes 5:95) to give the amide as colorless prisms (896
mg, 79%): mp 182 °C; [R]20D ) +43.6° (c 1.508, EtOAc).
(1S,2S,4R)-2-exo-(2′-Aminophenyl)-1,3,3-tri-
methylbicyclo[2.2.1]heptan-2-ol (11b). (a) (1′S,2′S,4′R)-
N-[2-(2′-Hydroxy-1′,3′,3′-trimethylbicyclo[2.2.1]hept-2′-
yl)phenyl]trifluoroacetamide. 2-Bromo-N-trifluoroaceta-
nilide (3.06 g, 0.011 mol) in THF (20 mL) cooled to 0 °C was
treated with methyllithium (1.3 M in diethyl ether, 9.0 mL,
0.012 mol) to give a colorless solution. The solution was cooled
to -80 °C, and tert-butyllithium (1.3 M in pentane, 17.6 mL,
0.023 mol) was added dropwise over 30 min so as to maintain
the temperature below -70 °C. The yellow solution was
stirred at -80 °C for 1 h and allowed to warm to -30 °C,
(26) Perrin, D. D.; Armarego, W. L. F. K.; Perrin, D. R. Purification
of Laboratory Chemicals; Pergamon: Oxford, 1980.
(27) Winkle, M. R.; Lansinger, J. M.; Ronald, R. C. J. Chem. Soc.,
Chem. Commun. 1980, 87.
(28) Kofron, W. G.; Baclawski, L. M. J. Org. Chem. 1976, 41, 1879.
(29) SPARTAN 4.0, Wavefunction, Inc., 18401 Von Karman, Suite
370, Irvine, CA 92715.
Rearrangement of 2-exo-Arylfenchyl Alcohols
becoming deep red in color. The resulting solution was treated
with fenchone at room temperature for 20 h (overnight). The
crude product was purified by flash chromatography (ethyl
acetate/hexane 10:90) to yield the pure hydroxy amide as
colorless prisms (2.79 g, 72%): mp 128-131 °C, [R]20D ) +9.8°
(c 1.852, EtOAc).
(b) (1S,2S,4R)-2-exo-(2′-Aminophenyl)-1,3,3-trimethyl-
bicyclo[2.2.1]heptan-2-ol (11b). A solution of the trifluo-
roacetamide (3.79 g, 0.011 mol) and sodium hydroxide (12.0
g, 0.30 mol) in ethanol (100 mL) was heated at reflux under
nitrogen for 24 h. The ethanol was removed under reduced
pressure, the residue was suspended in water, and the product
was extracted with ether. The combined ether layers were
dried (K2CO3), and the solvent was removed under reduced
pressure. The crude product was purified by flash chroma-
tography (ethyl acetate/hexane, 15:85) and further by distil-
lation [bp 150-160 °C/0.08 mm (Kugelrohr)] to yield the amino
alcohol 11b as colorless prisms (2.17 g, 80%): mp 54-55 °C,
[R]20D ) +181.3° (c 1.100, hexane).
(1S,2S,4R)-2-exo-[2′-(N,N-Dimethylamino)phenyl]-1,3,3-
trimethylbicyclo[2.2.1]heptan-2-ol (11c). A solution of
N,N-dimethylaniline (0.382 g, 3.16 mmol, 0.400 mL) and
TMEDA (0.459 g, 3.95 mmol, 0.459 mL) in diethyl ether (10
mL) was treated with butyllithium (1.88 mL, 3.94 mmol, 2.1
M) at room temperature for 30 min. The reaction mixture was
heated under reflux for 3 h and cooled to room temperature.
The lithiated aniline was treated with fenchone at room
temperature for 1.5 h and then for 30 min under reflux before
quenching with water. The crude product was distilled at 60-
70 °C (Kugelrohr) to remove the unreacted aniline and
fenchone, and the residue was purified by flash chromatog-
raphy (ether/hexane 10:90) to give the alcohol as a colorless
crystalline solid (0.262 g, 30%): mp 61-63 °C (lit.10 mp 57-
60 °C), [R]20D ) +69.2° (c 0.338, EtOAc).
(1′′S,2′′S,4′′R)-2-[2′-(2′′-Hydroxy-1′′,3′′,3′′-trimethyl-
bicyclo[2.2.1]hept-2′′-yl)phenyl]tetrahydropyran (11d). A
solution of 2′-(phenyloxy)tetrahydropyran (1.96 g, 0.011 mol)
in ether (10 mL) was treated with butyllithium (8.0 mL, 0.015
mol, 1.9 M) and the resulting slurry was stirred at room
temperature for 24 h. This was then treated with fenchone
at room temperature for 3 h. The crude product was purified
by flash chromatography (ethyl acetate/light petroleum 10:90)
to give the hydroxy THP ether as a colorless solid (2.10 g, 58%)
and as a 1:1 mixture of diastereomers: [R]20D ) +61.5° (c 1.524,
EtOAc).
(1′S,2′S,4′R)-2-(2′-endo-Hydroxy-1′,3′,3′-trimethyl-
bicyclo[2.2.1]hept-2′-yl)phenol (11e). A solution of the THP
ether 11d (0.130 g; 0.389 mmol) in THF (8 mL), water (4 mL),
and acetic acid (17 M, 1.6 mL) was heated under reflux for 2
h, cooled, and then poured onto water. The mixture was
extracted with ether, and the combined extracts were washed
with water and brine and dried (Na2SO4). The crude product
after evaporation of the solvent was purified by flash chroma-
tography (ether/hexane 15:85) to give the hydroxy phenol as
a colorless unstable crystalline solid (86 mg, 89%), mp 96-
100 °C, [R]20D ) +110.1° (c 0.345, EtOAc), which was used
immediately in subsequent experiments.
(1S,2S,4R)-2-exo-(2′-Methoxyphenyl)-1,3,3-trimethyl-
bicyclo[2.2.1]heptan-2-ol (11f). A solution of anisole (2.52
g, 0.023 mol) in diethyl ether (25 mL) was treated with
butyllithium (2.0 M, 14.0 mL, 0.028 mol), and then the reaction
mixture was heated at reflux for 22 h. The reaction mixture
was treated with fenchone at room temperature for 3 h. The
crude product was purified by flash chromatography (ethyl
acetate/hexane 5:95) to afford the alcohol as a colorless oil that
later solidified as colorless needles (3.85 g, 64%): mp 64-67
°C (lit.13 mp 65-66 °C); [R]20D ) +119.2° (c 0.302, EtOAc) (lit.13
[R]20D ) +118.46°).
(1S,2S,4R)-2-exo-(2′,6′-Dimethoxyphenyl)-1,3,3-tri-
methylbicyclo[2.2.1]heptan-2-ol (11g). A solution of 1,3-
dimethoxybenzene (3.04 g, 0.022 mol) in THF (30 mL) was
treated with butyllithium (13.2 mL, 2.0 M, 0.026 mol), and
then the reaction mixture was heated at 40 °C for 2 h. The
reaction mixture was treated with fenchone at room temper-
ature for 2 h. The unreacted 1,3-dimethoxybenzene and excess
J. Org. Chem., Vol. 63, No. 7, 1998 2269
fenchone were removed by distillation at 100 °C/0.2 mm
(Kugelrohr), and the resulting solid was recrystallized from
hexane to give the alcohol as colorless needles (4.37 g, 68%);
mp 73 °C, [R]20D ) +81.0° (c 0.790, EtOAc).
(1S,2S,4R)-2-exo-(4′-Methoxyphenyl)-1,3,3-tri-
methylbicyclo[2.2.1]heptan-2-ol (11h). A solution of p-
bromoanisole (0.50 g; 2.67 mmol) in diethyl ether (20 mL) was
treated with butyllithium (1.59 mL, 3.34 mmol, 2.1 M) slowly
at -5 °C. The reaction mixture was stirred at between -5
and 0 °C for 1 h before being treated with fenchone at 0 °C for
5 min and then at room temperature for a further 1 h. Flash
chromatography (ether/hexane 5:95) afforded the alcohol as a
colorless crystalline solid (0.529 g, 76%): mp 49-51 °C, [R]20D
) +43.2° (c 1.51, EtOAc).
(1S,2S,4R)-2-exo-[2′-(Hydroxymethyl)phenyl]-1,3,3-
trimethylbicyclo[2.2.1]heptan-2-ol (11i). A solution of
benzyl alcohol (0.523 g, 4.83 mmol, 0.500 mL) and TMEDA
(1.28 g, 0.011 mol, 1.28 mL) in dry hexane (15 mL) cooled in
an ice bath was treated with butyllithium (5.06 mL, 0.011 mol,
2.17 M) to give a pale precipitate. The reaction mixture was
heated under reflux for 4 h and cooled, and then diethyl ether
(10 mL) was added. This was then treated with fenchone at
room temperature for 60 h. Unreacted benzyl alcohol and
fenchone were removed by Kugelrohr distillation (70 °C/1 mm)
to leave a crystalline residue that was recrystallized from
hexane to give the alcohol as acid-sensitive colorless fluffy
needles (0.403 g, 32%): mp 115-116 °C, [R]20D ) +133.8° (c
0.208, EtOAc).
(1S,2S,4R)-2-exo-[2′-(Methoxymethyl)phenyl]-1,3,3-
trimethylbicyclo[2.2.1]heptan-2-ol (11j). A solution of the
diol 11i (96.6 mg, 0.368 mmol) and methyl iodide (52.3 mg,
0.368 mmol, 23 (L) in THF (5 mL) was treated with sodium
hydride (60% dispersion in oil) portionwise (5 × 5 mg) at room
temperature over a period of 18 h, after which time TLC
indicated that the reaction was complete. The reaction was
quenched with ammonium chloride solution, and the product
was extracted into ether. Flash chromatography (ether/hexane
10:90) of the crude product gave the monomethyl ether as a
colorless viscous oil (98.9 mg, 98%): [R]20D ) +128.8° (c 0.396,
EtOAc).
(1S,2S,4R)-2-exo-(2′-Methylphenyl)-1,3,3-trimethyl-
bicyclo[2.2.1]heptan-2-ol (11k). A solution of o-bromotolu-
ene (0.711 g, 4.16 mmol, 0.500 mL) in diethyl ether (25 mL)
was treated with butyllithium (2.47 mL, 5.20 mmol, 2.1 M)
dropwise at -10 °C, whereupon the resulting solution was
stirred for 1 h at -10 °C and then at 10 °C for 40 min. The
reaction mixture was treated with fenchone at 0 °C, and then
the whole was stirred at room temperature for 14 h, after
which time a fine precipitate had developed. Flash chroma-
tography (ether/hexane 5:95) afforded the alcohol as a viscous
colorless oil (0.825 g, 81%): [R]20D ) +109.9° (c 0.626, EtOAc).
Acid Hydrolysis Reactions. Method A. In EtOH/10 M
HCl. The arylfenchyl alcohol (0.03-0.05 M) in EtOH/10 M
HCl (1:1 or 2:1, v/v) was heated at the temperature indicated
in Table 2, and except in the cases of 11a, 11b, and 11c the
reaction progress was monitored by TLC. Upon completion,
after the reaction time given in Table 2, the reaction mixture
was cooled and poured onto water and the product extracted
into ether.
Method B. With TfOH in CHCl3. The arylfenchyl alcohol
(0.05-0.07 M in CHCl3) was treated with TfOH (1 equiv) as a
neat liquid dropwise with cooling in ice and under a nitrogen
atmosphere. The reaction mixture was then stirred at the
temperature indicated in Table 2. When TLC showed that
reaction was complete (time given in Table 2), the mixture was
poured onto water and extracted with ether.
(a) (i) 11a in EtOH/10 M HCl (1:1, v/v), 24 h. The
pivalamide (0.121 g; 0.333 mmol) was hydrolyzed according
to method A (Table 2, entry 1). Prior to ether extraction, the
reaction mixture was allowed to cool and treated with a
solution of KOH (4 g) in water (30 mL). The product mixture
was submittted to flash chromatography on a column protected
from light with aluminum foil (ether/hexane 10:90) to give as
the first fraction (2R,4aR,9aR)-9-aza-6-chloro-2,4a-(10,10-di-
methylmethano)-1,2,3,4,4a,9a-hexahydro-9a-methyl-9H-fluo-
2270 J. Org. Chem., Vol. 63, No. 7, 1998
rene (3a) as colorless plates (26 mg, 29%): mp 134-136 °C,
[R]20D ) -165.5° (c 0.220, EtOAc); IR (CHCl3) 3388m cm-1; 1H
NMR (400 MHz) δ 0.934 (3H, s, CH3), 0.946 (3H, s, CH3), 1.180
(3H, s, CH3), 1.278 (1H, ddd, J ) 12.3, 8.6, 5.3 Hz), 1.366 (1H,
d, J ) 12.5 Hz), 1.689 (1H, ddd, J ) 9.0, 9.0, 2.2 Hz), 1.84-
1.94 (2H, m), 1.97-2.04 (1H, m), 2.451 (1H, ddd, J ) 12.5,
2.9, 2.9 Hz), 3.523 (1H, br s, Wh/2 ) 23 Hz, NH), 6.467 (1H, d,
J ) 8.1 Hz), 6.944 (1H, d, J ) 2.0 Hz), 6.981 (1H, dd, J ) 8.2,
2.2 Hz); 13C NMR (50 MHz) δ 20.40, 20.94, 23.58 (CH2) 26.14
(CH2), 27.16, 44.50 (CH2), 46.37, 53.00 (quaternary), 64.00
(quaternary) 73.22 (quaternary), 110.55, 122.63 (quaternary),
123.64, 127.22, 130.27 (quaternary), 150.36 (quaternary); MS
m/z 261 (M, 40); HRMS m/z calcd for C16H20NCl M+ 261.1284,
found M+ 261.1253. Anal. Calcd for C16H20NCl: C, 73.41; H,
7.70; N, 5.35. Found: C, 73.40; H, 7.70; N, 5.66.
Next was eluted (1R,2S,4R,6S)-2-(2′-amino-5′-chlorophenyl)-
1,3,3-trimethyltricyclo[2.2.1.02,6]heptane (12a) as a colorless
crystalline solid (41 mg, 45%) and as a 73:27 mixture of
atropisomers: IR (CHCl3) 3497w, 3406w cm-1; 1H NMR (200
MHz) δ (*denotes minor atropisomer) 0.791* (3H, s, CH3),
0.838 (3H, s, CH3), 0.878 (3H, s, CH3), 0.984* (3H, s, CH3),
1.086 (3H, s, CH3), 1.130 (1H, br s), 1.237* (1H, br s), 1.285*
(1H, br s), 1.352* (1H, br d, J ) 11.1 Hz), 1.451 (1H, br d, J )
10.8 Hz), 1.598 (1H, br s, both isomers), 1.826 (1H, br d, J )
10.7 Hz), 1.86-1.93* (2H, m), 1.971 (1H, br d, J ) 10.7 Hz),
3.838 (2H, br s, Wh/2 ) 36 Hz, NH2), 6.55-6.61 (1H, m), 6.90-
7.08 (2H, m); 13C NMR (50 MHz) δ (*denotes minor atropiso-
mer) 14.68*, 15.40, 21.63, 21.77, 22.05 (quaternary), 22.57*,
24.73, 26.0 (quaternary), 26.31, 26.38*, 31.83* (CH2), 32.58
(CH2), 37.63 (CH2), 38.42* (CH2), 41.95, 43.19*, 48.41 (qua-
ternary), 48.90* (quaternary), 116.30*, 116.60, 121.87* (qua-
ternary), 121.98 (quaternary), 127.13, 127.27*, 133.10, 134.23*,
146.12* (quaternary), 146.40 (quaternary); MS m/z 261 (M,
75); HRMS m/z calcd for C16H20NCl M+ 261.1284, found M+
261.1282.
(ii) 11a in EtOH/10 M HCl (1:1, v/v), 48 h. Treatment of
11a (0.152 g, 0.416 mmol) for 48 h resulted in formation of
the hexahydroazafluorene 3a (68.5 mg, 63%).
(b) 11b in EtOH/10 M HCl (1:1, v/v), 24 h. Hydrolysis of
the amino alcohol 11b (78.9 mg, 0.322 mmol) (method A, Table
2, entry 2) and workup as in (a) afforded a solid residue that
upon purification by flash chromatography on a column
protected from light with aluminum foil (ether/hexane 3:97)
gave (2R,4aR,9aR)-9-aza-2,4a-(10,10-dimethylmethano)-1,2,3,
4,4a,9a-hexahydro-9a-methyl-9H-fluorene (3b) as colorless
prisms (50 mg, 68%): mp 64-66 °C; [R]20D ) -248.8° (c 0.118,
EtOAc) and as the only product: IR (CHCl3) 3384m cm-1; 1H
NMR (400 MHz) δ 0.936 (3H, s, CH3), 1.196 (3H, s, CH3), 1.288
(1H, ddd, J ) 12.5, 8.9, 5.5 Hz), 1.373 (1H, d, J ) 12.5 Hz),
1.707 (1H, ddd, J ) 12.3, 9.2, 2.8 Hz), 1.84-1.94 (2H, m), 2.01-
2.09 (1H, m), 2.465 (1H, ddd, J ) 12.5, 3.2, 3.2 Hz), 6.581 (1H,
d, J ) 7.7 Hz), 6.725 (1H, dd, J ) 7.3, 0.9 Hz), 7.00-7.06 (2H,
m); 13C NMR (100 MHz) δ 20.46, 21.03, 23.61 (CH2), 26.23
(CH2), 27.29, 44.63 (CH2), 46.41, 52.79 (quaternary), 63.93
(quaternary), 72.53 (quaternary), 109.74, 118.08, 123.44,
127.43, 128.33 (quaternary), 151.97 (quaternary); MS m/z 227
(M, 18); MS (CI methane) m/z 228 (M + 1, 100); HRMS m/z
calcd for C16H21N M+ 227.1674, found M+ 227.1673. Anal.
Calcd for C16H21N: C, 84.53; H, 9.31; N, 6.16. Found: C, 84.41;
H, 9.53; N, 5.96.
(c) 11c in EtOH/10 M HCl (2:1, v/v), 6 h. Hydrolysis of
the dimethylamino alcohol 11c (58 mg, 0.213 mmol) (method
A, Table 2, entry 3) and workup as above returned 11c that
by NMR analysis was unchanged, suggesting that dehydration
had not occurred.
(d) 11d in EtOH/10 M HCl (2:1, v/v), 20 min. The THP
ether 11d (97.3 mg, 0.292 mmol) was treated according to
method A (Table 2, entry 4). TLC indicated immediate
hydrolysis to the diol 11e. Workup, followed by flash chro-
matography (ether/hexane 1:99), afforded (3R,4aR,9bR)-3,9b-
(10,10-dimethylmethano)-1,2,3,4,4a,9b-hexahydro-4a-meth-
yldibenzofuran (3c) and 4a in a ratio of 95:5 as a colorless
viscous oil (58.7 mg, 88%), [R]20D ) -52.5° (c 0.282, EtOAc).
The product ratio was determined by integration of the signals
at 2.646 and 2.224 ppm in the 1H NMR spectrum. 3c: IR
Starling et al.
(CHCl3) 1614m, 1233s, 1160s, 1084s cm-1; 1H NMR (400 MHz)
δ 0.763 (3H, s, CH3), 0.904 (3H, s, CH3), 1.287 (3H, s, CH3),
1.26-1.34 (1H, m), 1.549 (1H, d, J ) 13.2 Hz), 1.629 (1H, ddd,
J ) 12.4, 8.8, 2.8 Hz), 1.93-1.97 (2H, m), 2.04-2.11 (1H, m),
2.646 (1H, ddd, J ) 13.6, 2.8, 2.8 Hz), 6.729 (1H, br d, J ) 8
Hz), 6.854 (1H, ddd, J ) 7.2, 7.2, 0.8 Hz), 7.076 (1H, dd, J )
7.2, 1.6 Hz), 7.129 (1H, ddd, J ) 7.6, 7.6, 1.2 Hz); 13C NMR
(50 MHz) δ 19.97, 20.48, 22.57 (CH2), 24.30, 26.13 (CH2), 43.57
(CH2), 45.92, 52.27 (quaternary), 63.62 (quaternary), 97.64
(quaternary), 110.05, 119.89, 123.46, 128.08, 160.55 (quater-
nary); MS m/z 228 (M, 69); HRMS m/z calcd for C16H20O M+
228.1514, found M+ 228.1521. Data for 4a are given in section
(f) (i).
(e) (i) 11e in EtOH/10 M HCl (2:1, v/v). Hydrolysis of
the diol 11e (35.8 mg; 0.145 mmol) (method A, Table 2, entry
5) (bath 60 °C) for 1.5 h gave complete conversion into a 97:3
mixture of 3c and 4a (26.4 mg, 80%). The reaction was
extremely slow at room temperature.
When the diol 11e (31.0 mg, 0.126 mmol) was treated at 87
°C, the reaction was complete after 15 min, giving a 94:6
mixture of 3c and 4a. The product mixture thereby obtained
was resubmitted to the same reaction conditions, and after 7
h 3c and 4a were obtained as a 60:40 mixture (26.7 mg, 93%).
(ii) 11e with TfOH (1 equiv) in CHCl3. Hydrolysis of the
diol 11e (85.7 mg, 0.348 mmol) (method B, Table 2, entry 5)
afforded a 93:7 mixture of 3c and 4a (72.5 mg, 91%) as a
viscous oil.
(f) (i) 11f in EtOH/10 M HCl (2:1, v/v). Hydrolysis of the
o-anisyl alcohol 11f (0.143 g; 0.550 mmol) (method A, Table 2,
entry 6) at 85 °C (bath) for 15 min gave complete conversion
into the cyclofenchene 12b.13 This was obtained by flash
chromatography (hexane) as a viscous oil (81 mg, 65%) and
as a 52:48 mixture of atropisomers.
Treatment of 11f (58.5 mg, 0.225 mmol) in the same medium
at 85 °C led to clean formation of (1R,4S,4aR,9bR)-1,2,3,4,4a,9b-
hexahydro-1,4-methano-1,4a,9b-trimethyldibenzofuran (4a)13
after 7.5 h. 4a was obtained as colorless needles (44.3 mg,
86%): mp 74-75 °C; [R]20D ) -36.8° (c 0.473, EtOAc); 1H NMR
(400 MHz) δ 0.877 (1H, dddd, J ) 12.6, 9.1, 4.5, 2.6 Hz), 1.060
(1H, ddd, J ) 12.6, 12.6, 4.9 Hz), 1.180 (3H, s, CH3), 1.235
(3H, s, CH3), 1.279 (1H, dd, J ) 10.3, 1.6 Hz), 1.20-1.40 (m,
2H), 1.360 (3H, s, CH3), 1.366 (1H, dddd, J ) 12.7, 12.7, 4.5,
4.5 Hz), 1.60-1.71 (2H, m), 2.224 (1H, ddd, J ) 4.4, 1.4, 1.4
Hz), 6.727 (1H, ddd, J ) 8.0, 0.9, 0.5 Hz), 6.809 (1H, ddd, J )
7.4, 7.4, 1.0 Hz), 6.997 (1H, ddd, J ) 7.4, 1.5, 0.5 Hz), 7.116
(1H, ddd, J ) 8.0, 7.4, 1.5 Hz); 13C NMR (50 MHz) δ 17.84,
19.48, 21.89, 23.56, 34.01, 42.18, 49.18, 50.81 (quaternary),
55.63 (quaternary), 97.27 (quaternary), 108.89, 119.58, 123.51,
128.01, 133.68 (quaternary), 158.71 (quaternary). NMR data
are in agreement with those reported in the literature.13
(ii) 11f with TfOH (1 equiv) in CHCl3. Hydrolysis of 11f
(87.2 mg, 0.335 mmol) (method B, Table 2, entry 6) afforded
the cyclofenchene 12b (60.1 mg, 74%) as a viscous oil.
(g) (i) 11g in EtOH/10 M HCl (2:1, v/v), 85 °C. Hydrolysis
of the dimethoxyphenyl alcohol 11g (85 mg; 0.293 mmol)
(method A, Table 2, entry 7) afforded (1R,4S,4aR,9bR)-4a,9b-
dimethyl-1,2,3,4,4a,9b-hexahydro-1,4-methano-9-methoxydiben-
zofuran (4b) after flash chromatography (ether/hexane 1:99)
as colorless needles (66.6 mg, 88%): mp 58-61 °C; [R]20D ) -
1.1° (c 0.460, EtOAc); IR (CHCl3) 1596s, 1254s, 1099s, 1082s
cm-1; 1H NMR (400 MHz) δ 0.998 (1H, dddd, J ) 12.1, 7.0,
4.3, 2.6 Hz), 1.115 (1H, ddd, J ) 12.2, 12.2, 4.8 Hz), 1.237 (1H,
dd, J ) 10.4, 1.6 Hz), 1.280 (3H, s, CH3), 1.301 (3H, s, CH3),
1.355 (3H, s, CH3), 1.380 (1H, dddd, J ) 12.6, 12.6, 4.4, 4.4
Hz), 1.57-1.65 (2H, m), 2.158 (1H, br d, J ) 4.6 Hz), 3.779
(3H, s, CH3), 6.358 (1H, d, J ) 8.8 Hz), 6.379 (1H, d, J ) 8.2
Hz), 7.059 (1H, dd, J ) 8.2, 8.2 Hz); 13C NMR (50 MHz) δ 16.94,
18.66, 21.87, 23.54, 34.85, 42.47, 48.76, 50.93 (quaternary),
54.92, 56.61 (quaternary), 97.52 (quaternary), 102.33 (quater-
nary), 102.47 (quaternary), 120.12 (quaternary), 128.83, 157.40
(quaternary), 160.01 (quaternary); MS m/z 258 (M, 20). Anal.
Calcd for C17H22O2: C, 79.03; H, 8.58. Found: C, 78.73; H,
8.71.
Treatment of 11g as above at 20 °C resulted in slow
conversion into 4b over 6 days.
Rearrangement of 2-exo-Arylfenchyl Alcohols
(ii) 11g in 17 M AcOH/48% HBr (4:1, v/v). A solution of
11g (1.081 g, 3.72 mmol) in acetic acid (17 M, 40 mL) was
treated with hydrobromic acid (48%, 10 mL, 7.15 g, 0.088 mol),
and the resulting purple solution was heated at reflux for 24
h. The solution was diluted with water and extracted with
ether. The combined organic layers were washed with water
and saturated sodium hydrogen carbonate solution and dried
(Na2SO4), and the solvent was removed under reduced pres-
sure. The products were separated by flash chromatography
(ethyl acetate/light petroleum 2:98 grading to 10:90) to give
first 4b (320 mg, 33%) followed by (1R,4S,4aR,9bR)-
1,2,3,4,4a,9b-hexahydro-9-hydroxy-1,4-methano-1,4a,9b-trim-
ethyldibenzofuran (4c) (316 mg, 35%) as colorless needles: mp
191-192 °C; [R]20D ) -3.7° (c 0.380, EtOAc); IR (CHCl3)
3595m, 3329br m, 1604s, 1294s, 1022s cm-1; 1H NMR (400
MHz) δ 1.02-1.09 (1H, m), 1.157 (1H, ddd, J ) 12.4, 12.4, 5.1
Hz), 1.262 (1H, dd, J ) 10.3, 1.5 Hz), 1.318 (3H, s, CH3), 1.338
(3H, s, CH3), 1.366 (3H, s, CH3), 1.399 (1H, dddd, J ) 12.6,
12.6, 4.5, 4.5 Hz), 1.60-1.67 (2H, m), 2.179 (1H, br d, J ) 3.6
Hz), 4.72 (1H, br s, Wh/2 ) 59 Hz, OH), 6.188 (1H, dd, J ) 8.0,
0.6 Hz), 6.341 (1H, dd, J ) 8.0, 0.6 Hz), 6.952 (1H, dd, J )
8.0, 8.0 Hz); 13C NMR (100 MHz) δ 17.03, 18.73, 21.86, 23.55,
34.76, 42.46, 48.83, 50.96 (quaternary), 56.39 (quaternary),
97.73 (quaternary), 102.12, 107.22, 118.84 (quaternary), 128.83,
153.14 (quaternary), 160.68 (quaternary); MS m/z 244 (M, 40).
Anal. Calcd for C16H20O2: C, 78.65; H, 8.25. Found: C, 78.41;
H, 8.05.
(h) (i) 11h in EtOH/10 M HCl (2:1, v/v), Reflux. Hy-
drolysis of the p-anisyl alcohol 11h (0.235 g, 0.901 mmol)
(method A, Table 2, entry 8) and flash chromatography
(hexane) afforded the cyclofenchene (1R,2S,4R,6S)-2-(4′-meth-
oxyphenyl)-1,3,3-trimethyltricyclo[2.2.1.02,6]heptane (12c) as
a colorless mobile oil (0.124 g, 57%): [R]20D ) 0.0° (c 0.490,
EtOAc); IR (thin film) 2955vs, 1609s, 1515s, 1295s, 1244s,
1173s, 1040s cm-1; 1H NMR (400 MHz) δ 0.738 (3H, s, CH3),
0.875 (3H, s, CH3), 0.987 (1H, ddd, J ) 1.2, 1.2, 1.2 Hz), 1.069
(3H, s, CH3), 1.225 (1H, dd, J ) 10.6, 1.2 Hz), 1.337 (1H, ddd,
J ) 10.6, 1.2, 1.2 Hz), 1.568 (1H, ddd, J ) 1.5, 1.5, 1.5 Hz),
1.817 (1H, ddd, J ) 10.5, 1.5, 1.5 Hz), 1.884 (1H, br d, J )
10.5 Hz), 3.801 (3H, s, CH3), 6.82-6.85 (2H, m), 7.12-7.15 (2H,
m); 13C NMR (50 MHz) δ 15.22, 21.59, 25.19 (quaternary),
25.68, 32.46 (CH2), 38.14 (CH2), 40.21 (quaternary), 43.02,
46.79 (quaternary), 55.08 (OCH3), 113.13 (2 signals), 129.54
(quaternary), 132.62 (2 signals), 158.04 (quaternary); MS m/z
242 (M, 22); MS (CI methane) m/z 243 (M + 1, 100). Anal.
Calcd for C17H22O: C, 84.25; H, 9.15. Found: C, 84.19; H, 9.19.
(ii) 11h with TfOH (1 equiv) in CHCl3. Hydrolysis of
the p-anisyl alcohol 11h (76.7 mg; 0.295 mol) (method B, Table
2, entry 8) afforded (1R,4R)-7,7-dimethyl-1-(4′-methoxyphe-
nyl)-2-methylenebicyclo[2.2.1]heptane (13) after flash chro-
matography (ether/hexane 1:99) as a colorless oil (59.5 mg,
83%): [R]20D ) +103.8° (c 0.208, EtOAc); IR (thin film) 1611s,
1289s, 1248s, 1181s, 1041s cm-1; 1H NMR (400 MHz) δ 0.746
(3H, s, CH3), 0.976 (3H, s, CH3), 1.339 (1H, ddd, J ) 12.2, 9.4,
4.6 Hz), 1.558 (1H, ddd, J ) 12.2, 9.4, 4.0 Hz), 1.853 (1H, dd,
J ) 4.4, 4.4 Hz), 1.917 (1H, ddddd, J ) 12.0, 12.0, 3.8, 3.8, 3.8
Hz), 2.093 (1H, ddd, J ) 16.0, 2.4, 2.4 Hz), 2.526 (1H, ddd, J
) 11.9, 11.9, 4.4 Hz), 2.654 (1H, ddddd, J ) 16.0, 3.1, 3.1, 3.1,
3.1 Hz), 3.810 (1H, s, OCH3), 4.538 (1H, dd, J ) 3.0, 3.0 Hz),
4.796 (1H, dd, J ) 2.8, 2.8 Hz), 6.84-6.88 (2H, m), 7.28-7.32
(2H, m); 13C NMR (50 MHz) δ 19.14, 21.66, 27.39 (CH2), 33.06
(CH2), 38.06 (CH2), 43.02 (quaternary), 46.14, 49.91 (quater-
nary), 55.17 (OCH3), 105.00 (CH2), 112.83, 129.88, 131.63
(quaternary), 157.79 (quaternary), 158.92 (quaternary); MS
m/z 242 (M, 91); HRMS m/z calcd for C17H22O M+ 242.1671,
found M+ 242.1669.
(i) 11i in EtOH/10 M HCl (2:1, v/v), Reflux. Hydrolysis
of the o-(hydroxymethyl)phenyl alcohol 11i (71.2 mg, 0.271
mmol) (method A, Table 2, entry 9) and chromatography
(ether/hexane 1:99) afforded (1′S,2′R,4′R)-spiro[phthalan-2,2′-
(1′,3′,3′-trimethylbicyclo[2.2.1]heptane) (14) as a colorless oil
(65.3 mg, 98%): [R]20D ) +25.0° (c 0.838, EtOAc); IR (KBr)
1608w, 1057s cm-1; 1H NMR (400 MHz) δ 0.698 (3H, s, CH3),
0.766 (3H, s, CH3), 1.038 (3H, s, CH3), 1.148 (1H, ddd, J )
12.5, 12.5, 4.1 Hz), 1.289 (1H, dd, J ) 10.5, 1.4 Hz), 1.496 (1H,
J. Org. Chem., Vol. 63, No. 7, 1998 2271
dddd, J ) 12.5, 12.5, 5.4, 4.6 Hz), 1.752 (1H, ddd, J ) 4.1, 2.2,
2.2 Hz), 1.864 (1H, dddd, J ) 12.3, 9.2, 4.5, 2.6 Hz), 2.095 (1H,
dddd, J ) 10.4, 1.9, 1.9, 1.9 Hz), 2.193 (1H, dddd, J ) 11.8,
9.1, 5.3, 2.3 Hz), 4.888 (1H, d, J ) 12.2 Hz), 4.953 (1H, d, J )
12.2 Hz), 7.13-7.23 (3H, m), 7.28-7.32 (1H, m); 13C NMR (50
MHz) δ 17.69, 23.01, 25.20 (CH2), 29.42, 30.59 (CH2), 41.51
(CH2), 46.10 (quaternary), 48.59, 52.51 (quaternary), 71.57
(CH2), 98.85 (quaternary), 120.53, 125.16, 125.78, 126.62,
140.51 (quaternary), 141.65 (quaternary); MS m/z 242 (M, 25);
HRMS m/z calcd for C17H22O M+ 242.1671, found M+ 242.1678.
Anal. Calcd for C17H22O: C, 84.25; H, 9.15. Found: C, 84.20;
H, 9.17.
(j) 11j in EtOH/10 M HCl (2:1, v/v), Reflux. Hydrolysis
of the o-(methoxymethyl)phenyl alcohol 11j (41.7 mg, 0.152
mmol) (method A, Table 2, entry10) afforded the phthalan 14
as a colorless oil (36.8 mg, 99%).
(k) (i) 11k in EtOH/10 M HCl (2:1, v/v), Reflux. Hy-
drolysis of the o-tolyl alcohol 11k (45.6 mg; 0.187 mmol)
(method A, Table 2, entry 11) under reflux for 40 min led to
complete conversion into the cyclofenchene 12d. Chromatog-
raphy (hexane) afforded (1R,2S,4R,6S)-2-(2′-methylphenyl)-
1,3,3-trimethyltricyclo[2.2.1.02,6]heptane (12d) as a colorless
oil (38.3 mg, 91%): IR (CHCl3) 2959vs, 1290m cm-1; 1H NMR
(400 MHz) δ (*denotes minor atropisomer) 0.756* (3H, s, CH3),
0.821 (3H, s, CH3), 0.836* (1H, ddd, J ) 1.1, 1.1, 1.1 Hz), 0.861
(3H, s, CH3), 0.923* (3H, s, CH3), 1.054 (3H, s, CH3), 1.112
(1H, ddd, J ) 1.1, 1.1, 1.1 Hz), 1.192* (3H, s, CH3), 1.264 (1H,
dd, J ) 10.4, 1.2 Hz), 1.277* (1H, dd, J ) 10.6, 1.2 Hz), 1.366*
(1H, ddd, J ) 10.6, 1.3, 1.3 Hz), 1.434 (1H, ddd, J ) 10.6, 1.4,
1.4 Hz), 1.57-1.60 (1H, m, both isomers), 1.839 (1H, ddd, J )
10.6, 1.6, 1.6 Hz), 1.933* (1H, br d, J ) 9.4 Hz), 1.949* (1H,
ddd, J ) 10.7, 1.6, 1.6 Hz), 2.012 (1H, br d, J ) 10.6 Hz), 2.381*
(3H, s, CH3), 2.392 (3H, s, CH3), 7.06-7.26 (4H, m, both
isomers); 13C NMR (50 MHz) δ (*denotes minor atropisomer)
15.03*, 15.75, 20.43, 20.67*, 21.87, 22.07, 22.34*, 25.15
(quaternary), 25.96*, 26.44* (quaternary), 26.95, 32.02* (CH2),
32.68 (CH2), 37.95 (CH2), 38.06 (quaternary), 38.46* (CH2),
40.23* (quaternary), 42.27, 43.33*, 48.54 (quaternary), 48.98*
(quaternary), 124.66, 124.93*, 125.98, 126.09*, 130.15*, 130.59,
133.62, 134.68*, 134.92 (quaternary), 135.99* (quaternary),
139.45* (quaternary), 139.92 (quaternary); MS m/z 226 (M,
42); MS (CI methane) m/z 227 (M + 1, 100). Anal. Calcd for
C17H22: C, 90.20; H, 9.80. Found: C, 90.00; H, 9.93.
When the alcohol 11k (0.180 g, 0.738 mmol) was treated
under the same conditions for 3 d a product mixture (0.117 g)
was obtained after workup and chromatography (ether/hexane
1:99) that was found by HPLC analysis (Whatman Partisil 5
column, flow rate 1.5 mL/min, refractive index detector, hexane
as solvent) to consist of the cyclofenchene 12d (84%) (tR ) 4.25
min) and a mixture of 11 unidentified species (16%) (tR ) 5.1-
10.1 min).
(ii) 11k in EtOH/10 M HCl (1:1, v/v), Reflux. Hydrolysis
of 11k (0.111 g, 0.453 mmol) (method A, Table 2, entry 11) at
93 °C for 24 h and chromatography (ether/hexane 1:99)
afforded a product mixture (75.3 mg) that was found by HPLC
analysis to consist of the cyclofenchene (64%) and the same
mixture of 11 species as above (36%).
(iii) 11k with TfOH (1 equiv) in CHCl3. Hydrolysis of
11k (96 mg, 0.393 mol) (method B, Table 2, entry 11) for 10
min resulted in complete conversion into the cyclofenchene 12d
as indicated by TLC analysis. The reaction mixture was
heated under reflux for a total of 24 h and afforded after
workup and chromatography (hexane) the cyclofenchene 12d
(21.7 mg, 24%) and an inseparable mixture of products (35
mg, 39% based on mass of 12d).
Acknowledgment. We thank the Australian Re-
search Council for financial support.
Supporting Information Available: Characterization
data (IR, 1H and 13C NMR, MS, and elemental analysis or
HRMS) for compounds 11a-e,g-k and (1′S,2′S,4′R)-N-[2-(2′-
hydroxy-1′,3′,3′-trimethylbicyclo[2.2.1]hept-2′-yl)phenyl]tri-
2272 J. Org. Chem., Vol. 63, No. 7, 1998 Starling et al.

fluoroacetamide, 1H NMR data for compound 11f, tabulated
NMR data with assignments for compounds 3c, 4a, 13, and
14, NOE results for compounds 3c, 4a, and 14, 1H NMR
spectra for compounds 3c, 11e, 12a, and 13, and analytical
HPLC chromatogram for hydrolysate from 11k (17 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
JO972025V